Author(s)
Federico Bobbio, Michael Honig, Randall Berry, Rakesh Vohra, Thanh Nguyen, and Vijay Subramanian
Abstract
We study an optimal mechanism design problem for resource sharing between an incumbent noncommercial user and a commercial entrant. Motivated by applications such as wireless spectrum allocation and access to national resources, where noncommercial and commercial users compete for limited availability, the mechanism is subject to two key constraints. First, direct transfers cannot take place to compensate the incumbent for externalities caused by the activity of the commercial user. Instead, the incumbent reports an interference disutility for sharing to a regulator; the second constraint is that this disutility can only be verified via a costly inspection. The regulator optimizes total welfare by announcing a probability of shared resource assignment along with a probability of inspection. For two kinds of interference disutilities, we show that in the optimal mechanism, both allocation and inspection decisions are binary: The regulator either fully allows or prohibits sharing, and inspections occur only for certain reported values when sharing is denied. We further reformulate the problem as a knapsack model to characterize threshold extremal points.